Israeli technology companies are building technology that can help save lives. Their tech can help find bombs but it can also evaluate the state of vehicles. It can recognise the driver and their state. And it can quickly inform medical agencies of predicted injuries. These companies are using AI, data, machine learning, algorithms, software or sensors to protect drivers. Auto Futures talked to representatives from three start-ups who explain how their technology works.
In Israel at security checkpoints, bombs attached to vehicles were traditionally discovered by someone holding a mirror under them, which didn’t make sense to Ohad Hever, Co-founder and COO of UVeye. It led him to create a system that uses cameras, machine learning and computer-vision that “within seconds detects any anomaly in a vehicle.”
“We’ve already started with big automakers in Japan and Europe putting the tech in the end of the assembly line,” says Hever.
The vehicle is driven through an arch over a base unit surrounding the vehicle with cameras for a 360-degree view. Images combined with cloud data and software can identify scratches and dents as well as mechanical problems. UVeye is involved in major pilots with OEMs and announced a partnership with SKODA.
“We’ve installed UVeye in more than 50 locations all over the world for security with great results,” says Hever about the company that is also has a pilot in Israel for commercial fleets scanning buses for mechanical issues.
“Because we have a lot of data, we can track parts over time for preventative maintenance,” adds Hever. For vehicle rental, sharing and leasing companies UVEye can detect scratches as small as 2mm. The company is also working on additional ways to sense anomalies and will be available for aftermarket repair shops to inspect frames, tires, brakes, wheels, exhaust, springs and exteriors.
Three years ago, computer vision company, eyeSight Technologies saw the opportunity was huge for using vision-based solution in cars, says Inbal Toren, Senior Product Manager at eyeSight.
Previously, the company developed viewer analytics that determined when there were five people in a household in front of a TV. It found not only the age and genders of the viewers but also that only two would be engaged with television content with a camera-enabled device.
“EyeSight’s driver monitoring software includes the ability to identify the driver as well as visual attributions such as pupils, head position and other data to determine if the driver is paying attention,” says Toren. She noted that eyeSight developed its algorithms using 200 academic resources involving consciousness and attention to form its vision-based information. The software has a high level of accuracy and is performed using cutting edge computing that can use processors as low as a dual-core ARM processor with a camera not directly in front of the driver.
Another problem is when drivers have advanced driver assistance features they may rely too much on them.
“Our solution can detect if someone may be appearing to be looking at the road but not paying attention such as someone with a cognitive load in a daydreaming situation,” says Toren. After the distracted or drowsy state has been detected the car may make a warning sound or vibrate the seat, suggesting the driver go for coffee.
The company is also working on an aftermarket solution and with a major Chinese vehicle manufacturer.
“We can detect if a driver is smoking,” says Toren. In parts of China it is illegal for bus drivers to drive while smoking.
In the future, eyeSight software can be used to detect if the human driver is able to take over from autonomous driving. “It’s amazing, the response to demonstrations has been very positive,” says Toren, “We have a good cause. We are saving lives.”
MDGo to the Rescue
Even with all the technology that prevent car accidents when a crash or collision occurs new technology from MDGo can help emergency responders and health care professionals care for injured drivers and passengers.
“We use existing sensors such as airbag deployment, seat belts,” says Gilad Avrashi Co-founder and CTO of MDGo. “And data crash tests from sources such as NCAP, NHTSA and IIHS and fill in the gaps with our own crash tests.”
MDGo interprets the data into a medical and mechanical report that is delivered to automakers, emergency medical services and trauma centres. The report describes car damage and injury severity.
“A telephone call can take up to five minutes, while MDGo data can be available in eight seconds,” adds Avrashi. The data is helpful for dispatchers and for medical personnel to prepare when responding to the accident.
MDGo research suggests that as many as 44% of car crash fatalities could be prevented with real-time and injury severity medical data were available. MDGo can also save paperwork by providing data for insurance carriers.
Unfortunately, the MDGo system and other safety systems were not in service when MDGo’s CEO Dr Itay Bengad, MD, was riding his bicycle and was hit by a car driven by a drowsy driver.
“Dr Bengad still suffers from pains and some (hopefully temporary) disability in his shoulders” says Avrashi. But he got right back to work on MDGo after surgery.